26 Hits in 7.6 sec

CAKES: Channel-wise Automatic KErnel Shrinking for Efficient 3D Networks [article]

Qihang Yu, Yingwei Li, Jieru Mei, Yuyin Zhou, Alan L. Yuille
2020 arXiv   pre-print
In this paper, we propose Channel-wise Automatic KErnel Shrinking (CAKES), to enable efficient 3D learning by shrinking standard 3D convolutions into a set of economic operations e.g., 1D, 2D convolutions  ...  Further, we propose a new search space based on CAKES, so that the replacement configuration can be determined automatically for simplifying 3D networks.  ...  To address these issues, we introduce Channel-wise Automatic KErnel Shrinking (CAKES), as a general framework to automatically determine an efficient replacement for existing 3D operations.  ... 
arXiv:2003.12798v3 fatcat:hfpkixc5efdoxjgowwg7grczta

Searching for Two-Stream Models in Multivariate Space for Video Recognition [article]

Xinyu Gong, Heng Wang, Zheng Shou, Matt Feiszli, Zhangyang Wang, Zhicheng Yan
2021 arXiv   pre-print
In this work, we present a pragmatic neural architecture search approach, which is able to search for two-stream video models in giant spaces efficiently.  ...  We demonstrate two-stream models with significantly better performance can be automatically discovered in our design space.  ...  CAKES [46] searches channel-wise spatial/temporal kernels to improve model efficiency.  ... 
arXiv:2108.12957v1 fatcat:743coyuvjvf27lsgwrgxy56faq

Efficient Deep Learning: A Survey on Making Deep Learning Models Smaller, Faster, and Better [article]

Gaurav Menghani
2021 arXiv   pre-print
, and also equip them with ideas for further research and experimentation to achieve additional gains.  ...  We also present an experiment-based guide along with code, for practitioners to optimize their model training and deployment.  ...  Similarly there is an order of magnitude less computation since the point-wise convolution is much cheaper for convolving with each input channel depth-wise (for more calculations refer to [133] ).  ... 
arXiv:2106.08962v2 fatcat:33kuvqwrojc5tf3ozud5acz7ou

Bruker Micro-CT User Meeting 2016 Abstract book [article]

2020 Zenodo  
Acknowledgements We thank Prof Rob van 't Hof for access to XMT image data.  ...  Acknowledgements To Bruker-Skyscan staff for fast and effective support, their patience and effectiveness, and for their constant improvements to the software and in implementing new options we requested  ...  During the 3D printing stage, the scaffold density was changed in order to generate more or less dense micro-channels networks.  ... 
doi:10.5281/zenodo.3697293 fatcat:lcotsgt3wvdzxpiltd4zczg7sy

A Special issue on Multidisciplinary research to promote the applications in Biological sciences

Guest Editor Dr. M. Manikandan
2022 International Journal of Life Science and Pharma Research  
This provides a platform for the researchers to publish their innovative research findings.  ...  It also includes papers on phytochemical screening of medicinal plants and also focuses on wireless sensor networks in health care applications.  ...  Normally, Bayes Shrink, Visu Shrink, normal shrink and Sure Shrink are evaluated with this threshold function.  ... 
doi:10.22376/ijpbs/ijlpr/sp22/2022.1-154 fatcat:rcydgammlzgodof3kw3fcdgpji

EMG-based Simultaneous and Proportional Estimation of Wrist Kinematics and its Application in Intuitive Myoelectric Control for Unilateral transradial Amputees

Farina Dario
2011 Frontiers in Computational Neuroscience  
Neural Networks 21, 1070-1075. Acknowledgements We thank M. Kaczorowski for her assistance with the histological procedures.  ...  The financial support by the Interdisciplinary Center for Scientific Computing, the International Graduiertenkolleg 710 (DFG) and the funding by National Genome Research Network (a program of the German  ...  An automatic or semi-automatic method for song classification would speed up this tedious process.  ... 
doi:10.3389/conf.fncom.2011.53.00081 fatcat:nkq5wesfpbcjnikbxqh3v3gtqi


Somrita Chattopadhyay
And, finally, for the Massachusetts Buildings Dataset, our network achieves 96.19% relaxed IoU score [...]  ...  We also report improvements on the previous-best results for the challenging INRIA Validation Dataset for which our network achieves an overall IoU of 81.28% and an overall accuracy of 97.03%.  ...  The Critic network has 4 strided convolutional layers with kernel size 7 for the first 2 layers and kernel size 5 for the next 2 layers.  ... 
doi:10.25394/pgs.19322168 fatcat:ietap2bwxjcltb3u6jl2sgccsm

Mobile biomass gasification unit coupled to an internal combustion engine - Design, realization and investigation:

Δημήτριος Π. Μερτζής
Furthermore, the 1D kinetic model developed during this thesis for the prediction of the producer gas composition is compared to the actual findings regarding olive kernel gasification.  ...  The reactor is not only evaluated for its fluidization behaviour but also for its performance.  ...  The efficiency of collection increases as the depth of the filter cake increases, making these filters highly efficient for small diameter, even sub-micron particles size.  ... 
doi:10.26262/ fatcat:soq4ipdlj5c6xnadnvzhrmhxnm

Advanced functional materials

2000 Advanced Materials for Optics and Electronics  
for Life (RM4L)'.  ...  Hugh Bruck of University of Maryland for their guidance and support.  ...  Moreover, the 3D arrangement gives channels for particle relocation.  ... 
doi:10.1002/1099-0712(200001/02)10:1<1::aid-amo420>;2-p fatcat:y2u4dfrznje4lksromwcbtt3gy

2013 AACC International Annual Meeting Abstracts

2013 Cereal foods world  
Quinoa seeds (QS) were purchased from a local supermarket, germinated for 7 days in an automatic sprouter and maintained at 10°C under fluorescent light.  ...  Flour yield was significantly (P<0.05) different for large, medium, and small kernels across mills. Large kernels on Quadrumat.  ...  IWAR correlated positively by the percentage of pericarp (r=0.679) of the kernel, and negatively by kernel density (r=-0.671) and test weight (r=-0.602).  ... 
doi:10.1094/cfw-58-4-a fatcat:et4urgwq3bhb5g5y4k7zf2hkzy

Interactive Segmentation of 3D Medical Images with Implicit Surfaces

Benoit Mory
Despite continuing advances in mathematical models for automatic segmentation, many medical practitioners still rely on 2D manual delineation, due to the lack of intuitive semi-automatic tools in 3D.  ...  These properties are key factors for clinical acceptance.  ...  As a result, 3D applications can be envisaged without compromising efficiency.  ... 
doi:10.5075/epfl-thesis-5247 fatcat:q22b37atzzcqff7i6ihmjoaqai

LIPIcs, Volume 32, SNAPL'15, Complete Volume [article]

Thomas Ball, Rastislav Bodik, Shriram Krishnamurthi, Benjamin S. Lerner, Greg Morrisett
We thank Josef Widder for the fruitful discussions on this subject, Heather Miller and Philipp Haller for their help with the pickling library, and the anonymous reviewers for their comments.  ...  The Need for Language Support for Fault-Tolerant Distributed Systems Robin Milner, Joachim Parrow, and David Walker Toward a Dependability Case Language and Workflow for a Radiation Therapy System Acknowledgments  ...  ISA support for efficient bounds checking.  ... 
doi:10.4230/lipics.snapl.2015 fatcat:ljoxwa7wp5achn2tbrmpriez7y

13th International Bologna Conference on Magnetic Resonance in Porous Media - Bologna 2016 Conference Handbook and Book of Abstracts

Villiam Bortolotti, Paola Fantazzini, Leonardo Brizi
We also thank Françoise LE CAHEREC b , Marie-Françoise NIOGRET b , Carole DELEU b , Nathalie NESI b , Anne LAPERCHE b et Christine BISSUEL b for setting up the field experiment and for cooperation on this  ...  Acknowledgements ACKNOWLEDGEMENTS We thank the Regional Council of Bretagne for financial support.  ...  This makes wise setting of digital filters even more critical.  ... 
doi:10.6092/unibo/amsacta/5446 fatcat:nft5sqhwizdqli7ev6cfxehiwi

Development and characterization of a dispersion-encoded method for low-coherence interferometry [article]

Christopher Taudt
2021 arXiv   pre-print
The selection of a dispersive element for encoding allows for tuning of the axial measurement range and resolution of the setup.  ...  The approach is theoretically designed and implemented for applications such as surface profilometry, the characterization of polymeric cross-linking and as a tool for the determination of layer thicknesses  ...  I will always be thankful for your support!  ... 
arXiv:2105.10246v1 fatcat:xwo6yjrrdvdjhp5clxao3ewk2i

Dagstuhl Reports, Volume 6, Issue 6, June 2016, Complete Issue [article]

After reviewing some of the existing model based time series kernels, I will introduce a framework for building new kernels based on temporal filters inspired by a class of parametrized state space models  ...  Land-value data: For land, in contrast to cake, there exist detailed value maps, which can be used to test the performance of cake-cutting algorithms.  ...  It takes over control of the driving as much as the driver desires or needs, keeps the driver in the loop if necessary, communicates efficiently, explains in an appropriate way its status, behavior, intentions  ... 
doi:10.4230/dagrep.6.6 fatcat:bahw4gi2tjfdth3kvsi3k2uc6i
« Previous Showing results 1 — 15 out of 26 results